gopalakrishnan okstseparation of resin types in mixed bed ion exchange columnate 0664m 11405

SEPARATION OF RESIN TYPES IN MIXED BED

ION EXCHANGE COLUMNS

By

BHARATHWAJ GOPALAKRISHNAN

Bachelor of Science in Chemical Engineering

St. Joseph‟s College of Engineering

(Affiliated to Anna University)

Chennai, Tamil Nadu

2009

Submitted to the Faculty of the

Graduate College of the

Oklahoma State University

in partial fulfillment of

the requirements for

the Degree of

MASTER OF SCIENCE

May, 2011

ii

SEPARATION OF RESIN TYPES IN MIXED BED

ION EXCHANGE COLUMNS

Thesis Approved:

Dr. Gary L. Foutch

Thesis Adviser

Dr. Josh Ramsey

Dr. Heather Fahlenkamp

Dr. Mark E. Payton

Dean of the Graduate College

.

iii

ACKNOWLEDGMENTS

I wish to express my deepest gratitude to my advisor Dr. Gary L. Foutch. What I am today is

largely due to his inspiration, kindness, understanding and invaluable guidance throughout the

course of my masters program. I would like to thank Mr. William Moore from Aquatech who

trained me to work on the pilot plant. I would to thank my colleague Mr. Joon Yong Lee for his

assistance and help with my study. I am very grateful to Dr. Heather Fahlenkamp and Dr. Josh

Ramsey for serving on the advisory committee and aiding me with the completion of this study.

I am very thankful for the funding provided by Aquatech International Corporation and Anadarko

Petroleum Corp. Special thanks must be given to my parents S. Gopalakrishnan and R. Lakshmi

and my fiancé S. Pavithra whose emotional support was unfailing throughout my time at

Oklahoma State University. Their love and support deserve my undying gratitude.

.

iv

TABLE OF CONTENTS

Chapter Page

I. INTRODUCTION ......................................................................................................1

Ion exchange ............................................................................................................3

Mixed bed ion exchange ..........................................................................................3

Need for effective resin separation ..........................................................................4

II. RESIN SEPARATION IN MIXED BEDS ...............................................................7

Early development ...................................................................................................7

Condensate polishing ...............................................................................................8

Resin separation techniques ...................................................................................10

Conventional approach: Density separation ......................................................10

Interface isolation method .................................................................................11

Inert resins .........................................................................................................11

Conesep separation system ................................................................................13

Fullsep separation system ..................................................................................14

Seprex process ...................................................................................................16

Sepra eight process ............................................................................................17

Resin on resin approach ....................................................................................18

Resin cleaning, sizing and separation system ..................................................19

Amsep process ...................................................................................................20

Efficient and effective resin separation..................................................................22

III. GRAVITY SEPARATION OF RESINS ...............................................................24

Introduction ............................................................................................................24

Description of experimental equipment .................................................................24

Estimation of cross contamination .........................................................................25

Backwash results: H/OH form ...............................................................................26

Steady state bed expansion ................................................................................36

Effect of water temperature on cross contamination .........................................36

Backwash results: Na/SO4 form.............................................................................39

Resin transfer study................................................................................................46

Discussion ..............................................................................................................48

Conclusion and recommendations .........................................................................50

v

REFERENCES ............................................................................................................52

APPENDIX A – Resin separation pilot plant operating manual .................................55

APPENDIX B – Experimental data .............................................................................71

APPENDIX C – Bead rise and fall model ...................................................................80

vi

LIST OF TABLES

Table Page

1 Effect of temperature on steady state bed expansion .............................................38

2 Resin transfer study................................................................................................47

3 Summary of backwash conditions used .................................................................49

4 Estimation of Reynolds numbers from flow rates .................................................49

vii

LIST OF FIGURES

Figure Page

1 Predicted equilibrium leakage from DOWEX 650C and 550A ...............................5

2 Interface isolation method......................................................................................12

3 Conesep mixed bed separation system ..................................................................14

4 Fullsep separation system ......................................................................................15

5 Seprex process .......................................................................................................17

6 Resin cleaning sizing and separation system .........................................................20

7 Amsep separation process ......................................................................................22

8 Fraction of anionic beads in cation layer at 13.5 gpm and 100% expansion .........27

9 Fraction of cationic beads in anion layer at 13.5 gpm and 100% expansion .........27



12 Fraction of anionic beads in cation layer at 4 gpm and 50% expansion ..............29

13 Fraction of cationic beads in anion layer at 4 gpm and 50% expansion ..............29

14 Fraction of anionic beads in cation layer at 13.5 gpm and 150% expansion .......30

15 Fraction of cationic beads in anion layer at 13.5 gpm and 150% expansion .......30

16 Fraction of anionic beads in cation layer at 6.5 gpm and 75% expansion ...........31

17 Fraction of cationic beads in anion layer at 6.5 gpm and 75% expansion ...........31


viii


20 Differences in anionic fraction within cation layer at 13.5 gpm flow .................33

21 Differences in cationic fraction within anion layer at 13.5 gpm flow .................33

22 Differences in anionic fraction within cation layer at low flow ..........................34

23 Differences in cationic fraction within anion layer at low flow ..........................34

24 Differences in anionic fraction within cation layer at 100% expansion ..............35

25 Differences in cationic fraction within anion layer at 100% expansion ..............35

26 Differences in anionic fraction within cation layer at 18oC and 40

oC .................37

27 Differences in cationic fraction within anion layer at 18oC and 40

oC .................37

28 Comparison of cationic fraction at 13.5 gpm and 100% expansion ....................40

29 Comparison of anionic fraction at 13.5 gpm and 100% expansion .....................40

30 Comparison of cationic fraction at 13.5 gpm and 75% expansion ......................41

31 Comparison of anionic fraction at 13.5 gpm and 75% expansion .......................41

32 Comparison of cationic fraction at 13.5 gpm and 150% expansion ....................42

33 Comparison of anionic fraction at 13.5 gpm and 150% expansion .....................42

34 Comparison of cationic fraction at 4 gpm and 50% expansion ...........................43

35 Comparison of anionic fraction at 4 gpm and 50% expansion ............................43

36 Comparison of cationic fraction at 6.5 gpm and 75% expansion ........................44

37 Comparison of anionic fraction at 6.5 gpm and 75% expansion .........................44

38 Comparison of cationic fraction at 6.5 gpm and 100% expansion ......................45

39 Comparison of anionic fraction at 6.5 gpm and 100% expansion .......................45

40 Comparison of conditions used and DOWEX expansion data ............................48

41 Force balances for rise and fall of a resin bead ....................................................50

ix

NOMENCLATURE

v Velocity of the resin bead (m/s)

d Diameter of resin bead (m)

m Mass of the resin bead (g)

V

Volume of the resin bead (m3)

mf Mass of fluid contained in the same volume as the bead (g)

g Acceleration due to gravity (m/s2)

ρ Density of the resin bead (g/ml)

ρf Density of the fluid (g/ml)

Cd Drag coefficient

Ratio of fluid to bead density

1

CHAPTER I

INTRODUCTION

Water, as we know, is indispensable to human life, playing a multitude of roles helping

us sustain our day to day activities. Industries rely on water for all levels of production.

Water can be used as a raw material, coolant, solvent, energy source and as a transport

agent. Many modern industries require high purity water to ensure process continuity and

product quality. Water of extremely high purity (ultrapure water) is required by several

specific industries such as, coal and nuclear power plants, pharmaceutical and

semiconductor manufacturing.

Process water with ionic impurity less than 1 µg/kg (ppb), with corresponding low levels

of particulate and microbial contaminants, is termed as high purity or ultrapure water

(Sadler 1993). The impurities to be removed during its production include ionic,

microbial, dissolved gases, organics, colloids and particulates. Commercially available

techniques for reducing the concentration of ionic species in water include reverse

osmosis, electrodialysis, ultrafiltration and ion exchange. Ion exchange plays a vital role

in the manufacture of high purity water. The demand for purer water led to innovations in

ion exchangers and ion exchange techniques. Mixed bed ion exchange methods use high

quality synthetic resins which remove ionic impurities present in water.

2

The focus is now on optimization of ion exchange techniques resulting in increased

productivity. Hence, modern ion exchange processes need to meet the demands for higher

quality water at low cost and with minimum waste. Resin separation in mixed bed ion

exchange is a vital step during the regeneration process and determines the extent of

sodium and sulfate or chloride into the system from caustic and acid regenerants,

respectively. The objectives of this study are:

a) Identify the effect of two variables: backwash flowrate and bed expansion on the

cross contamination of resin types during separation.

b) Compare cross contamination for two different forms of resin, the H/OH form and

Na/SO4 form.

c) Investigate the efficiency of interface isolation methods for resin transfer utilizing a

pilot scale plant.

3

ION EXCHANGE

Ion exchange is a branch of separation science which deals with the partition of charged

species between different system regions (Townsend 1993). Ion-exchange is defined as

the reversible stoichiometric interchange of ions between a solid phase (the ion-

exchanger) and a solution; the ion exchanger is usually insoluble in the medium in which

exchange occurs. Ion exchange is now a firmly established unit operation and an

extremely valuable alternative to other operations such as adsorption, distillation and

filtration. Though used in several chemical processes, ion exchange finds applications in

three broad categories: removal of ions, substitution and separation. Ion exchange can

effectively remove sodium, chloride, calcium, magnesium and other ions to produce large

volumes of ultrapure water required by food, power, microelectronics and pharmaceutical

industries.

MIXED BED ION EXCHANGE

In a mixed bed ion exchange column, cation and anion resins are adjacent through

mixture. In H-OH form the cation resin contains exchangeable H+ ions and anion resin

contains exchangeable OH- ions. Consequently, even better purification is obtained than

with the application of an alternating sequence of cation and anion exchangers because

exchanges are coupled by water dissociation equilibrium of the H+

and OH- ions to water

(Dorfner 1972). Hence, a mixed bed ion exchange column represents a powerful single

unit for demineralizing water; and with refinement, produces water of utmost purity.

4

NEED FOR EFFECTIVE RESIN SEPARATION

As water is continuously passed through the mixed bed, the cation and anion resins

remove the impure ions and release equivalent ions back into the water. In a normal H-

OH form operation the cation resin releases H+ ions and the anion exchanger releases

OH- ions. Eventually the ion exchange resin will saturate and their ability to exchange

ions gradually decreases and leads to bed exhaustion. When this condition is reached the

resins have to be regenerated to their original forms so that they can be used again. Anion

resin is regenerated with caustic (NaOH) and cation resin is regenerated with a suitable

mineral acid such as sulfuric or hydrochloric acid.

The most important step in mixed bed regeneration is resin separation. Improper

separation leads to cross contamination of the resin. Some of the cation resin, for

example, will contaminate the anion resin layer. Since the anion resin is subsequently

regenerated with caustic, the entrapped cation resin in the anion resin is completely

saturated with sodium ions. Similarly, when some of the anion resin contaminates the

cation resin, the entrapped anion resin is completely saturated with sulfate or chloride

ions depending on the acid used for cation resin regeneration.

Cross-contamination of resins leads inevitably to equilibrium leakage of sodium and

sulfur/chloride ions into the treated water. Figure 1 shows initial equilibrium leakage

from a mixed bed as a function of the initial ionic loading (Grimshaw and Harland 1975).

The initial loading is the fraction of sites in the sodium or chloride form on the cationic or

anionic resins, respectively.

5

Figure 1. Predicted equilibrium leakage from Dowex 650C and 550A

(Yi and Foutch 2004)

This could be from beads caught in the other resin prior to regeneration, or regeneration

efficiency as a result of insufficient chemical treatment or poor rinsing. The chloride

curve is lower in equilibrium leakage than the sodium curve because of the higher anionic

selectivity coefficient for chloride to hydroxide on the resin. Figure 1 indicates that if 1%

of the exchange sites are in the sodium form (99% in the hydrogen form) then the

concentration from the bed initially will be about 0.01 parts-per-billion. For 0.1% cross-

contamination the initial sodium leakage from the service cycle should be about 1.0 parts-

per-trillion. For chloride a 1% cross-contamination of anionic beads in the cation treated

with hydrochloric acid regenerate will yield an initial leakage of about 1.1 parts-per-

trillion. If sulfuric acid is the cationic regenerant, a 1% cross-contamination would result

in significantly lower leakage because of the higher preference of divalent ions by the

anionic exchange resin. This ignores the sulfate that may originate from potential cationic

degradation.

6

For ordinary deionization, minute contamination is of little consequence. However, for

applications like condensate polishing where one or two parts per billion (ppb) can be

serious, cross contamination of resins cannot be tolerated. Mixed bed resins operated in

the ammonia form leak 100 to 1000 times higher levels of impurities than the same resins

operated in the hydrogen cycle (Webb and Larkin 1996). Hence the target for allowable

cross contamination in ammonium form mixed bed is as low as 0.08% (Sadler 2001).

Therefore, if the purity of water from mixed beds is to be maintained at ultra-low

contaminant concentrations it is essential that the ion exchange resins are separated

completely.

7

CHAPTER II

RESIN SEPARATION IN MIXED BEDS

EARLY DEVELOPMENT

The first recorded use of mixed cation and anion exchangers consisted of weakly acidic

and basic resins that were non-regenerable (Martin 1952). The first use of strongly acidic

and basic exchangers was reported by McGarvey and Kunin (1951) and was called a

mono bed as the initial aim was to replace the two bed deionization with a single bed

system. The anion resin was separated from the cation resin and transferred to a separate

vessel for regeneration. The cation resin was regenerated in the service vessel. After

regeneration the anion resin was again transferred back to the service vessel. A second

patent, in 1954, described a mixed bed process where regeneration of both types of resin

took place within the service vessel (McGarvey and Kunin 1954).

It was soon found that the concept of single mixed bed replacing a two bed system

required more frequent regeneration and higher operating costs. There were problems

with precipitation of insoluble metal hydroxides during regeneration. These issues led to

the approach of using a two bed cation-anion exchange system followed by a mixed bed.

8

The mixed bed removed the trace impurities leaking from the two bed system. It received

a relatively small ionic load and had less frequent regeneration. Majority of the plants

then adopted the cation-anion exchange plus the mixed bed principle, although some

installations with high quality raw water adopted a mixed bed only approach.

CONDENSATE POLISHING

Throughout the power, utility and chemical industries, large volumes of water are

evaporated and condensed to either separate the solvent from solute, act as a driving force

for turbines and for many other purposes. In most cases the condensates are recycled.

Although these condensates are distilled waters of relatively high purity, in many

instances they must be purified before they can be reused. Feed water specifications of a

modern advanced high pressure steam raising plant are so low that direct re-use of

condensed steam from turbines is unacceptable. The condensate which may be

contaminated with cooling water and corrosion products from the steam generator, the

turbine, the preheater or the pipelines as a result of leaks in the tube bundles, must be

purified before it is used again. Hence the condensate needs to be demineralized or

'polished' by mixed bed ion exchange techniques employing resins. This technique is

mainly advantageous since it permits very high flow rates.

The water purity required from boiler feed water make-up plants is a conductivity of <

0.15-0.20 µS/cm and silicate concentrations of < 0.02 ppm (Dorfner 1972). Most of these

plants employ mixed bed systems containing a mixture of strong acid cation resin and

strong base anion resin. The resin is in service vessels which treat the condensate at a

9

relatively high specific flow rate as higher rates improve filtration and reduces the

number of vessels and amount of resin required.

Because of the high condensate flow rates, condensate polishing plants operate at much

higher throughputs than normal make up water treatment units. At such high flow rates,

the kinetics of ion exchange resins gives satisfactory operation but as the resins age, their

performance deteriorates. This leads to organic foulants in the mixed bed that affects

performance. The resins used today in condensate polishing are typically produced from

styrenedivinylbenzene (SDVB) copolymers. The SDVB copolymer consists of

polystyrene chains with divinylbenzene cross linking between the styrene chains. The

SDVB copolymer based ion exchange resin has the advantage over most other synthetic

polymers from a stability and capacity standpoint (Webb and Larkin 1996). Present day

manufacturing techniques allow physically stronger resins to be produced with tailored

and very consistent particle sizes. There is now a widespread use of Uniform Particle

Size (UPS) resins in condensate polishing with benefits being seen in the separation

process (Shields et al. 2006) and, consequently, in the quality of the final water.

Condensate polishing requires properly designed equipment, properly selected resin and

well trained operators. A comprehensive surveillance and monitoring program is also

required to maximize potential benefits of the whole operation. Condensate polishing has

found extensive application especially in the nuclear and power industry, where there is

demand for high purity water. Although advances have been made to produce greater

water quality there still remains a need to economize the process and increase efficiency.

An important step to achieve this high efficiency is the effective separation of resins prior

to regeneration.

10

RESIN SEPARATION TECHNIQUES

A wide range of methods are available for the effective separation of anion and cation

resin prior to regeneration. The methods vary from simple to complicated processes and

are chosen depending on the separation desired, economics, and spatial constraints.

CONVENTIONAL APPROACH: DENSITY SEPARATION

The majority of mixed bed separation processes available rely on the slightly different

falling speeds of anion and cation resins in water. Resin mixture in the separation vessel

is backwashed into suspension and allowed to settle. The denser cation resin beads settle

more quickly than the anion resin beads. An effective separation is generally achieved

with an anion resin upper layer and a cation resin lower layer. The interface consists of a

mixture of both cation and anion resin beads. The top anion layer is transferred

hydraulically via a side take-off port to an anion regeneration vessel. Resins are

regenerated separately, returned and mixed before rinsing. This method was initially

developed by McGarvey and Kunin after which improvements were done to the whole

process design for better resin separation and transfer (McGarvey and Kunin 1951).

Another conventional method was to regenerate both the cation an anion resin in the

same vessel after the density separation. A regenerant collector/distributor is located at

the interface for respective regenerants (McGarvey and Kunin 1954) however, it is

seldom used today. A well operated and maintained conventional regeneration system

may not achieve much better than 2 - 5 volume % cation in anion resin and 5 - 10 volume

% anion in cation resin after separation and transfer (Sadler 1986). However, with

11

customized operating procedures, it is possible to reduce cross-contamination to 0.4%

anion in cation resin and 0.5% cation in anion resin (Crone 1987).

INTERFACE ISOLATION METHOD

The interface isolation approach was a non-proprietary technique developed in the early

1970‟s. In this method, the anion and cation are separated by their density similar to

initial steps of the conventional approach. As shown in Figure 2, after separation, the

anion resin is transferred hydraulically to an anion regeneration tank. The interfacial layer

containing a mixture of both resins is removed to an isolation vessel via a second take-off

set just below the interface whilst the pure cation resin remains in the vessel (Hattori

1989), (Kusano and Nawata 1987).

The separated cation and anion resins are then regenerated using the appropriate

regenerants. This significantly minimizes cross contamination of anion in cation and

cation in anion. The interface is isolated till the next batch of spent resin arrives for

regeneration. This system requires control over the height of interface, that is, over the

resin volumes in each charge.

INERT RESINS

The use of inert resins having an intermediate density between that of anion and cation

resins was introduced in the year 1976 (Shields et al. 2006). These inert resins formed a

distinct buffer layer between the upper anion resin layer and the lower cation resin layer.

This minimizes the cross contamination of anion and cation resins at the interface. These

beads need to be ionically neutral and chemically and physically inert to the classifying

fluids and the regenerants.

12

Figure 2. Interface isolation method (Kusano and Nawata 1987)

The inert beads also had to impart a different conductivity than the anion and cation

beads so that the conductivity sensors used in certain separation systems could detect

them (Lefevre and Sato 1981). Soon after its advent, inert beads with enhanced qualities

such as better stability, wettability, uniformity, better strength and hydraulic stability

were developed (Chonde 1985; Osei-Gyimah 1988).

Recently, as field experience was gained and better separation methods developed,

interest in these resins decreased. They are known to foul and attract bubbles, causing

problems during handling and transfer. The advent of Uniform Particle Size resins, which

allowed excellent separation, also led to the phasing out of inert resins.

13

CONESEP SEPARATION SYSTEM

This high efficiency mixed bed separation system was first described in 1977 with a full

scale plant at Aghada, ESB, Ireland in 1980. The system consists of a vessel for resin

separation/anion regeneration, a cation regeneration vessel and a resin isolation vessel as

shown in Figure 3. Exhausted mixed bed resins are initially transferred to the resin

separation/anion regeneration vessel which has a conical bottom section. They are

cleaned and separated by backwashing in the usual manner. The lower cation resin layer

is transferred to the cation regeneration vessel via a transfer pipe drawing from the base

of the conical bottom section. To aid this transfer, a flow of water is introduced through

the porous base of the Conesep unit with part of this flow bled from the top of the vessel

and the remainder acting as resin motive water. The slight upflow helps to keep the resins

fluidized, stabilizing the cation/anion interface. As the cation resin is transferred, the

cation/anion interface gradually descends through the vessel into the conical section and

eventually enters the transfer pipe without any disturbance. A sharp resin interface passes

along the transfer pipe where it is detected by sensors, either by conductivity or optical

methods, and transfer into the cation regeneration vessel is stopped. The small quantities

of cation and anion resin held in the transfer line then directed to the interface isolation

vessel and combined with the next exhausted mixed bed charge entering the Conesep

system for regeneration. To achieve the highest separation efficiency, the anion resin is

regenerated with sodium hydroxide and then given a second separation by backwashing.

This takes advantage of the fact that any cation resin entrained in the anion resin will now

be in the denser sodium form and the anion resins in less dense hydroxide form.

14

Figure 3. Conesep mixed bed separation system (Emmet and Grainger 1979)

The very small volume of cation resin thus recovered is again directed to the interface

isolation vessel. Cation in anion resin cross contamination levels of about 0.05 to 0.07

volume % are usually obtained (EPRI 1997). A significant advantage of this method is

that the bottom transfer systems do not depend on the separated cation/anion interface

being at a precise height in the vessel. Thus operators have the freedom to adjust resin

ratios to suit their needs (Emmet 1977).

FULLSEP SEPARATION SYSTEM

Southern California Edison Company San Onofre Nuclear Generating Station developed

a condensate polishing system similar to the Conesep process during 1983 (Shields et al.

15

2006). The system is described in Figure 4. A cation-mixed bed arrangement was chosen

with an innovative design for separation/regeneration.

Figure 4. Fullsep separation system (Auerswald 1987)

It employs both interface isolation and bottom transfer of cation resin in the same process

(Auerswald 1983). The system was first commissioned in September, 1985. The San

Onofre mixed bed separation system has proved to be very effective, yielding both low

16

cation in anion resin levels as well as low anion in cation resin levels (Auerswald 1987).

The design was adopted commercially and has been supplied to many power stations.

SEPREX PROCESS

This proprietary separation improvement technique, also called the caustic floatation

process, was developed by the Graver Water Division of the Graver Company (EPRI

1997). In this method, the exhausted cation and anion resin are initially separated into

two distinct layers by backwashing similar to the conventional separation system. The

anion resin layer is drawn off and transferred to the anion regeneration vessel along with

the cation resin present in the interfacial layer. The anion resin along with the cationic

contaminants is regenerated with concentrated solution of sodium hydroxide (16 - 20 %

by weight).

The solution has a density such that the small volume of entrained cation resin sinks to

the bottom of the vessel while the anion resin floats on the top. The floated, regenerated

anion resin is transferred to the resin storage vessel with the caustic solution and rinsed

before combining with regenerated cation resin. The sodium form cation resin in the

anion regeneration vessel is transferred to the cation regeneration vessel where it is

regenerated with the next batch of resin. A schematic representation of Seprex process is

shown in figure 5. The removal of most of the cross contaminating cation resin leads to a

useful reduction in the overall sodium levels in the final mixed bed (Olijar and Salem

1970).

17

CR AR RS

1) Exhausted resin bed and Na form resin from AR vessel transferred to CR vessel for regeneration.

2) Resins cleaned, backwashed and hydraulically separated. Anion resin moves to top with cation resin in the bottom.

3) Anion resin, including mixed resin at interface, transferred to AR vessel.

4) Cation resin regenerated and rinsed. Anion resin floated and regenerated.

5) Floating anion resin transferred to RS vessel leaving Na form cation in AR vessel. Anion resin rinsed in RS vessel.

6) Cation resin transferred to RS vessel. Resin mixed and rinsed ready for use.

H2O

NH35-10%

H2SO4

16-20%NaOH

CR – Cation resin regeneration vesselAR – Anion resin regeneration vesselRS – Resin storage vessel

CR AR RS

Figure 5. Seprex process (Webb and Larkin 1996)

SEPRA EIGHT PROCESS

This system, introduced in 1983, combines the techniques employed in the Seprex and

Conesep processes (Salem and Scheerer 1983). System operation through the bottom

transfer of cation resin to the cation regeneration vessel follows the Conesep procedures.

However, inert resin is also used in the Conesep separation vessel to minimize anion

18

resin contamination of the cation resin. The cation resin is regenerated in the

conventional manner with 4 -10% by weight sulfuric acid. The anion, inert and small

amount of entrained cation resin are left in the separation vessel.

The anion resin is regenerated and floated to the anion resin rinse vessel as in the Seprex

process. After rinsing the regenerated cation and anion resins, the anion resin is

transferred to the cation regeneration vessel where the conventional air mixing and mixed

bed rinse are performed.

RESIN ON RESIN APPROACH

This procedure originated in South Africa in the mid 1970‟s (Sadler and Bolton 1996). It

was originally designed to reduce the problem of enhanced sodium leakage arising from

the level of cross contamination from conventional mixed bed separation systems where

top anion resin layer is drawn off to a separate vessel. However, it is now applied to

systems where cation resin is separated and transferred. Resins are first separated as

effectively as possible and the anion and cation resin is transferred to their respective

regeneration vessel. The anion resin, containing traces of cation resin, is regenerated and

rinsed the normal way. Any entrained cation resin is fully converted to the sodium form

in this process. The regenerated anion resin, containing the entrained cation resin, is

transferred back to the cation regeneration vessel and air mixed with the cation resin that

is still awaiting regeneration. The remixed resin is allowed to stand with periodic air

mixing then separated. The cation resin is regenerated and rinsed in the normal way

before being recombined with the anion resin for a final rinse before service.

19

This non-proprietary procedure reduces the sodium contamination of the final mixed bed.

Its use requires that the regeneration plant has the necessary connections to allow the

separated and regenerated anion resin to be remixed with the cation resin. Since its

advent, this method has been modified and customized to suit the particular location and

to increase separation efficiency. For example, the resin on resin approach coupled with

the interface isolation method showed significant improvement in separation efficiency

(Sadler 1995). One fossil power station estimates that it reduces the sodium originating

from cross contamination by a factor of 10 thus correcting for less than perfect separation

(McCarthy and O‟Connor 1992). The procedure has been adopted by many PWRs to

allow them to achieve very low sodium leakage when operating in the H-OH form.

Reports from these stations are positive with a typical comment being that it has allowed

them to reduce their sodium values to an acceptable level with essentially zero cost with

the exception of a bit of time.

RESIN CLEANING, SIZING AND SEPARATION SYSTEM

In this procedure, there is a steady feed of resins to be cleaned, sized and separated. As

shown in Figure 6, the exhausted mixed bed resin feed is initially passed through a pre-

treatment section where crud and other impurities are removed. The resin is fed into a

column of water flowing upwards in a specially designed vessel. The flow of water in this

vessel is carefully regulated in such a way that the anion resins alone are swept upwards.

The heavier cation resins sink to the bottom of the vessel. The anion resins overflow into

the anion resin settling section and into an anion regeneration vessel while the cation

resins are moved to a cation regeneration vessel. The remaining cation resin is transferred

to the cation regeneration vessel.

20

Condensatedemin

A

Condensatedemin

B

Condensatedemin

C

Resin receiving

tank

Resin separation/transfer process

Return to service vessels

SERVICE VESSELS

Resin transfer water

Anion regen tank

Cation regen tank

Resin storage

tank

Cation resin

Anion resin

Transfer water to drain

Figure 6. Resin cleaning, sizing and separation system (Shields et al. 2006)

AMSEP PROCESS

The Amsep process is based on the density differences between anion and cation resin in

the sulfate and ammonium (or other amine) forms, respectively. The first step in this

process is to air scrub and backwash the resin. This removes some iron from the surface

of the resin and facilitates resin separation to upper anion and lower cation. It prevents

magnesium precipitation from occurring within or on the anion resin during ammonium

sulfate regeneration.

21

The second step is to regenerate the separated resin using 2% ammonium sulfate as SO4.

This regeneration completes the conversion of cation into the ammonium form and

removes the exchanged calcium, magnesium and sodium. It also begins the conversion of

the anion resin into the sulfate form. Following the 2% ammonium sulfate regeneration

step, 6% ammonium sulfate is introduced into the resin in an upward flow. This step

completes the conversion of the anion into the sulfate form and generally results in the

liberation of carbon dioxide from the anion resin. The upward flow during this step

separates the cation and anion resins while not affected by the liberation of carbon

dioxide gas. By introducing the ammonium sulfate in steps, efficient and effective

stripping of chloride and sodium from the resin is accomplished. The final step is to float

the anion resin out of the separation tank and into the anion regeneration tank using 30%

ammonium sulfate. Ammonium sulfate is transferred from the anion regeneration tank to

the ammonium sulfate measuring tank thereby minimizing its consumption. 30%

ammonium sulfate solution has a density intermediate the cation and anion resins

allowing this flotation to occur readily at 60oF (EPRI 1997). At higher temperatures, a

slightly higher concentration of ammonium sulfate is needed to float the anion resin.

Following transfer, the cation resin is sluiced to the cation regeneration tank and both

resins are rinsed, air scrubbed and backwashed. The cation resin is regenerated with

sulfuric acid and the anion resin with caustic followed by a short ammonia rinse. This

process offers an advantage of cleaning the resin beads thoroughly in addition to the near

complete separation achieved. Reports of its performance at the supercritical Mohave

Power Station show that the Amsep process is very effective (Auerswald and Cutler

1991; Meyers 1993).

22

Ammonium sulfate

measuring tank

(30%)

40% bulk ammonium sulfate

storage

Anion regen tank

Separation tank

30% ammonium sulfate

Figure 7. Amsep separation process (Webb and Larkin 1996)

EFFICIENT AND EFFECTIVE RESIN SEPARATION

Choosing the right separation method depends on the purity of water desired, spatial

availability, manpower and economic constraints. Production of high purity water needs

greater separation efficiency which requires use of advanced separation techniques.

Operators need to be well trained to perform the process effectively. Customization and

modification of processes are sometimes necessary to meet spatial and economic

limitations. Conventional regeneration systems with appropriate customization can

reduce cross contamination to about 0.4% anion in cation resin and 0.5% cation in anion

resin (Crone 1987). Utilization of inert resin beads can reduce cross contamination of

23

cation in anion resin from 0.2 – 2% depending on the care with which separations are

performed (Bevan 1989). The Conesep system can achieve cross contamination levels of

<0.1% cation in anion resin and <0.5% anion in cation resin (Sadler 1986). A “tall thin”

separation tank with interface tank can limit cross contamination to 0.03% cation in anion

and 0.15% anion in cation (Auerswald 1986). Utilization of Resin cleaning and sizing

system can reduce resin cross contamination to 1 – 2% range for both phases of separated

resins (Stengel and Pillow 1990). The Amsep process can achieve a separation efficiency

of 0.01% cation in anion and 0.01% anion in cation (Auerswald and Cutler 1991).

24

CHAPTER III

GRAVITY SEPARATION OF RESINS

INTRODUCTION

Irrespective of the resin separation technique used prior to regeneration, a mixed bed has

to be differentiated into two distinct layers of anion and cation resin before transfer. This

is achieved by the gravity separation method which makes use of repeated backwash and

settling steps. Resin separation is directly proportional to the degree of fluidization of the

bed. The degree of fluidization can be affected by the velocity of the backwash water,

duration of flow and bed expansion during isothermal conditions. During backwashing,

the DI water enters from the bottom of the service vessel at a set flow rate thereby

fluidizing the resin bed and causing it to expand. Although there is a sharp difference in

particle density between the two types of resin, there still remain small quantities of resin

intermixing both below and above the interface. It is these quantities of cross-

contamination that will be evaluated in this study.

DESCRIPTION OF EXPERIMENTAL EQUIPMENT

The experimental unit is a full system containing three 12 inch columns, a large feed tank

and all required plumbing for multiple operations. Water can be fed from either the top or

25

bottom of each vessel and water and resin can be transferred among all three columns.

Air can be fed to all three vessels as required for specific operations. The service column

has six sampling valves spaced six inches apart that allow both sampling and resin

transfer above and below an interface. A 440 V pump moves water through the system.

Sensors for conductivity and pH are connected to sampling lines that can be moved to

multiple locations. The DI water feed tank includes a heater to operate the system at any

steady state temperature and a level controller to prevent equipment problems. A more

detailed description of the experimental unit can be found in appendix A. The Cation

resin used was DOWEX 650C and the anion resin was DOWEX 550A in a 2:1 volume

ratio.

ESTIMATION OF CROSS CONTAMINATION

For each trial a backwash flow rate and volume expansion percentage were selected and

five backwashing steps were performed. At low flow rates it is also possible to operate

for an extended time at the maximum bed expansion. However, higher flow rates can

result in moving resin out of the top of the column. Initially an air mix is used to

homogenize the bed prior to backwash. Ten minutes of air mixing was used prior to each

experiment. After each backwash, the interface height was recorded and water was

drained from the bed to allow it to stabilize within the column. Based on the geometry

and the amount of resin present, the samples were obtained from the ports immediately

above and below the interface. The samples were drawn from the take-off ports and then

processed to estimate the weight fraction of anion resin in the cation layer and cation

resin in the anion layer. Samples were placed in 20% NaCl where beads separate by

density difference. This step converts the anion and cation beads to the chloride and

26

sodium forms, respectively. Anion beads were siphoned off. All resins were placed on

filter paper, drained, rinsed and dried. Rinsing ensured that free salt was removed that

might interfere with an accurate weight of small masses. The filter paper and resin were

weighed, the resin removed and the filter paper reweighed and subtracted. Any salt

accumulated on the filters was included in the tare weight so that the weight recorded was

only for whole beads.

BACKWASH RESULTS: H/OH FORM

The initial set of experiments was performed with the cation resin in hydrogen form and

the anion resin in the hydroxide form. Eight experiments were repeated with a fixed

flowrate and bed expansion height. The first two experiments provided a chance to

practice operating the equipment and fine tune the experimental procedure. The flowrate

used for the first two experiments was 27 gpm (100% rotameter range) with a bed

expansion of 100%. This was too high a flow rate for optimum performance and resulted

in high cross contamination. The next flowrate chosen was 13.5 gpm (50% rotameter)

and the bed was allowed to expand to 100%.

The results for 13.5 gpm flow and 100% bed expansion are shown in Figures 8 and 9.

From these data the anionic beads in the cation layer controls the separation process since

the fraction of anionic resin decreased with each of the five successive washes. Error bars

were added considering a maximum deviation of 0.001 g during measurement. The error

bars are prominently visible for the anion layer due to relatively better separation

achieved from the very first backwash. Cationic resin beads in the anionic phase did not

indicate a correlation with the number of washes. This could be due to some cationic

27

beads being pushed up along with the with anion resin as a result of high flow. However,

all values were less than 0.35%, even after the first wash. A single wash showed 7%

anionic resin in cation layer. The third wash resulted in less than 1% anionic beads. The

fifth wash showed 4 cationic beads in an 8.667 g sample. The Dowex MSDS sheet

indicates that a backwash rate of 8 gpm should be sufficient for a 100% anionic bed

volume expansion. The observed values agreed with this assessment since increased

turbulence was clearly noted at higher flowrates.

Figure 8. Fraction of anionic beads in cation layer at 13.5 gpm and 100% expansion

Figure 9. Fraction of cationic beads in anion layer at 13.5 gpm and 100% expansion

28

The next experiment used a flowrate of 6.5 gpm and 100% bed expansion and the results

are shown in Figures 10 and 11. The fraction of anionic resin in the cation is lower

overall as compared to Figures 8 and 9, although the fifth backwash is not as low,

indicating the variance in the experimental data or the abilities to sample the column.

Results indicate that 0.2% can be achieved after the third wash and maintained through

subsequent washes. The cation in anionic resin range was only slightly improved with

three of five washes achieved less than 0.15% while only one wash achieved this level at

the higher flow rate.



29

Figures 12 and 13 are results obtained at 4 gpm and 50% bed expansion. The fraction of

anionic resin in the cation is significantly improved. The low flow and duration of

expansion was not sufficient to effectively raise the anion beads from the cation layer.

However, the fraction of cationic resin in the anion shows that four out of five

backwashes have less than 0.3% cross contamination.

Figure 12. Fraction of anionic beads in cation layer at 4 gpm and 50% expansion

Figure 13. Fraction of cationic beads in anion layer at 4 gpm and 50% expansion

30

Figures 14 and 15 are results obtained at 13.5 gpm and 150% bed expansion. The results

for anion layer weren‟t as good as expected. To achieve a greater bed expansion, a

higher flow rate is required which in turn has been noted to cause more turbulence

between the anion and interface layer. The minimum flow needed to achieve 150%

expansion was observed to be 13.5 gpm. A higher flow rate produces significantly more

mixing that lifts the cationic beads higher into the vessel. When flow is stopped the 50%

additional settling height is not sufficient to compensate for the increased mixing effect.



31

Figures 16 and 17 present results obtained at 6.5 gpm and 75% expansion. A flowrate of

4 gpm was not sufficient for 75% bed expansion. After five backwashes, the fraction of

cation resin in anion was 0.04% and the anion fraction in cation was 0.2%. There is

steady decrease in cross contamination after repeated backwash at the same conditions.

The data indicate that greater bed expansion gives better resin separation at the same flow

rate. The cation in anion contamination is significantly low, whereas the anion in cation

contamination is higher than that for 100% backwash expansion with the same flowrate.



32

Figures 18 and 19 are results obtained at 13.5 gpm and 75% expansion. The higher

flowrate causes increased mixing at the interface resulting in variation of cross

contamination values with successive backwashes. The bed expansion is not sufficient to

compensate for the increased mixing.



In order to more easily visualize comparisons with the data in Figures 8 through 19,

combined charts were prepared and are presented. The three types of comparisons

33

available from these data are: high flow (13.5 gpm) with different backwash expansion;

low flow (6.5 gpm and 4 gpm) with different backwash expansion; and 100% expansion

at different flows (13.5 gpm and 6.5 gpm). Figures 20 and 21 give the results for three

different expansions at a flowrate of 13.5 gpm. Figure 20 does show improvement in the

anionic resin within the cationic resin layer, particularly for the first two backwashes.

However, during the same experiments the amount of cationic resin in the anionic

fraction is marginally worse; as indicated in Figure 21.

Figure 20. Differences in anionic fraction within cation layer at 13.5 gpm flow

Figure 21. Differences in cationic fraction within anion layer at 13.5 gpm flow

34

Comparison of three different bed expansions (50%, 75% and 100%) at low flow rates

are shown in figures 22 and 23. Figure 22 shows a clear preference for 100% backwash

for anionic separation from the cation layer.

Figure 22. Differences in anionic fraction within cation layer at low flow

Figure 23. Differences in cationic fraction within anion layer at low flow

For the effects of the backwash flowrate with the same bed expansion (Figures 24 and

25) we can see a preference for the lower flow for the anionic resin fraction in the cation

resin layer. The low flow experiment gave the lowest measureable anionic resin value of

35

0.1%. For cationic resin in the anion layer the average improvement at the lower flow is

about 10%, with 3 of 5 values less than 0.15%.

Figure 24. Differences in fraction of anionic resin in the cation layer with flow for 100%

expansion.

Figure 25. Differences in fraction of cationic resin in the anion layer with flow for 100%

expansion.

36

STEADY STATE BED EXPANSION

In this experiment the bed was allowed to expand to its maximum height for a set flow

rate and the steady state expansion was maintained for a fixed period of time. The bed

was then allowed to settle and samples were drawn to measure the cross contamination in

each layers. The intent was to define what duration of steady-state expansion was

equivalent to performing 5 backwashes with the same flowrate. A flow rate of 4 gpm was

chosen for this experiment. The bed expanded by 58% and the steady state expanded

height could be maintained for 2 minutes 30 seconds. The fraction of cation in anion

resin was found to be 0.27% and the fraction of anion in cation was 0.78%. After five

backwashes at 4 gpm and 50% expansion, the fraction of cation in anion was 0.27%,

similar to the steady state expansion; whereas the fraction of anion to cation was 0.23%

which is significantly lower than the steady state value. The disadvantage of using the

steady state expansion method for separation is that lower flow rates do not push out the

anion resins from the cation layer effectively, whereas higher flowrates result in lesser

steady state expansion times.

EFFECT OF WATER TEMPERATURE ON CROSS CONTAMINATION

Viscosity and density of water changes when there is a change in water temperature. This

change should affect the rise and settling of resin beads. In order to investigate effect of

temperature change an experiment was conducted with the DI water at 40oC and

compared with results from normal experimental conditions (18oC). From the NIST data

for properties of liquid water, it is observed that the dynamic viscosity of water decreases

37

by 34% and density decreases by 0.6% when the temperature increases from 18oC to

40oC. The results for experiments at 75% expansion and flowrate of 6.5 gpm for both

temperatures are shown in Figures 26 and 27. The fraction of anionic resin in cation after

fifth backwash at 40oC was 67% lower than at 18

oC. This indicates lesser resistance

experienced by the resin beads due to elevated water temperature which pushes out the

lighter anion resin from the cation layer easily during backwash. The fraction of cation

resin in anion layer, however, remains the same for both experiments at 0.05%.

Figure 26. Differences in fraction of anionic resin in the cation layer at 18oC and 40

oC

Figure 27. Differences in fraction of cationic resin in the anion layer at 18oC and 40

oC

38

For separation of anion resin from the cation layer, low values of cross contaminations

are noticed for the first two backwashes itself. This could be due to the faster settling of

cation beads from the anion layer due to lesser resistance as the bed is allowed to settle.

A steady state experiment was also carried out at 40oC at a flow rate of 4 gpm. The bed

expanded by 44% which was lesser compared to 58% at 18oC. The expanded steady state

could be maintained for 2.15 minutes. As shown in Table 1, the fraction of anionic resin

in cation layer is much higher at 40oC. This is due to the reduced bed expansion at

elevated temperature. The fraction of cationic resin in anion layer, however, is lower than

at 18oC due to the ease of settling during the steady expanded state. Since only one test

was performed, results should be considered inconclusive.

Table 1. Effect of temperature on steady state bed expansion

Temperature (oC) Resin fraction

Cation layer Anion layer 40 0.0137 0.0007 18 0.0078 0.0027

39

BACKWASH RESULTS: Na/SO4 FORM

With continuous use of a mixed bed ion exchange column, the resins are slowly depleted

of H+ and OH

- ions. Ion exchange of impurities in water will eventually exhaust the resin

capacity sufficiently that resin regeneration is required. In order to compare the changes

in separation efficiency between fully loaded resins and resins in their pure form, gravity

separation experiments were repeated with the cation and anion resins converted to

sodium and sulfate forms, respectively. The two factors that affect separation in the

converted form are density and resin bead size. Cation resin shrinks by 6-7% and anion

resin by 13-15% in their converted forms (Applebaum 1968). The shrinkage is

accompanied by increase in density. The effect of these physical changes on the cross

contamination after separation was investigated.

The entire volume of cation resin was converted to sodium form by soaking the resin in

sodium hydroxide solution containing an equivalent amount of sodium ions for an hour.

Similarly the anion resin was soaked in sulfuric acid solution containing an equivalent

amount of sulfate ions. The resin beds were rinsed with DI water until any excess ions

were removed from the beds. Both resins were returned to the same column after

conversion. All six separation experiments with the resins in H/OH forms were repeated

in the converted forms.

Figures 28 and 29 show the comparison of cross contamination between H/OH and

Na/SO4 form resin for a flow rate of 13.5 gpm and 100% expansion. After five

backwashes, the converted form has lower cation cross contamination of 0.07% than the

0.3% observed in the H/OH form. However, for the anion layer, only two of five

40

backwashes were less than 0.5% for the converted form resin, while the H/OH form had

less than 0.5% cation fraction for all five backwashes. The cation layer has more anion in

the Na/SO4 form for all five backwashes. Increased turbulence was noticed in cation layer

for Na/SO4 form resin at a flow rate for 13.5 gpm. Though cross contamination gradually

decreased with each backwash, Na/SO4 form had an anion fraction of 1.5% in the cation

layer; which was ten times higher than the value in H/OH form.

Figure 28. Comparison of cationic fraction at 13.5 gpm and 100% expansion

Figure 29. Comparison of anionic fraction at 13.5 gpm and 100% expansion

41

Results for 13.5 gpm and 75% bed expansion are shown in Figures 30 and 31. The

anionic layer cross contamination values are erratic for both forms of resin due to high

flow and reduced expansion. However, at the end of five backwashes, Na/SO4 form

shows lower cross contamination of 0.09% compared to H/OH form (1.32%). The error

bars here are not big enough to explain a specific trend. So, there is more error than

measurement alone anticipated. The effect of increased mixing is also noticed in the

cation layer results for Na/SO4 form. Na/SO4 form has a higher anionic fraction of 2.3%

compared to 1.8% in H/OH form.



42

Results comparing both forms of resin at 13.5 gpm and 150% bed expansion are shown in

Figures 32 and 33. Results for anion layer indicate that after five backwashes, cationic

fraction in anion layer for Na/SO4 form resin has a slightly higher value of 0.097%

compared to H/OH form (0.063%). Results for cation layer show that lower anionic

fraction of 0.014% is observed in Na/SO4 form after five backwashes compared to

0.023% for H/OH form. This could be due to the increased density of cation resin in

Na/SO4 form resulting in decreased mixing in the cation layer at high bed expansion.



43

Results comparing both forms of resin at a low flow rate of 4 gpm and 50% bed

expansion are shown in Figures 34 and 35. The results for anion layer indicates

fluctuation with each backwash for H/OH form whereas a slightly uniform decrease in

cross contamination is noted in the Na/SO4 form. The final value of cationic fraction is

lower for Na/SO4 form (0.08%) compared to H/OH form (0.2%). Cation layer results

show that Na/SO4 form resin has a lower anionic fraction of 0.13% after five backwashes

compared to H/OH form which has 0.23%.

Figure 34. Comparison of cationic fraction at 4 gpm and 50% expansion

Figure 35. Comparison of anionic fraction at 4 gpm and 50% expansion

44

Figures 36 and 37 show results comparing both forms of resin at a flow rate of 6.5 gpm

and 75% bed expansion. The anion layer has higher cationic fraction of 0.5% in Na/SO4

form; 12 times greater than that in H/OH form after five backwashes. Results for cation

layer show close values for both forms of resin with 0.23% of anionic fraction for Na/SO4

form and 0.26% for H-OH form.



45

Results comparing both forms of resin at a flow rate of 6.5 gpm and 100% bed expansion

are shown in Figures 38 and 39. The anion layer shows cationic fractions greater 1% for

all five backwashes for Na/SO4 form. The increased bed expansion does not compensate

for the mixing effect between the interface and anion layer. The results for cation layer,

indicates that the anionic fraction is lower for Na/SO4 form (0.07%) compared to H/OH

form (0.1%) effectively removing the anion resin from the cation layer.

Figure 38. Comparison of cationic fraction 6.5 gpm and 100% expansion


46

RESIN TRANSFER STUDY

After separation of resins, efficient removal of anion layer to the anion regeneration

vessel and interface layer to the resin storage vessel had to be confirmed. The objective

was to verify if interface isolation method was to effectively maintain the purity of both

layers during resin transfer.

An initial test was done to fine tune the transfer process. After the gravity separation of

cation and anion layers, the CR vessel was completely filled with water. The take-off port

near the anion layer was connected to the anion regeneration column through a transfer

tube. The column was pressurized to 15 psig before transfer. No visible cross

contamination was noted in the anion layer during transfer. After transferring the anion

layer, the take-off port near the interface layer was connected to the resin storage vessel.

A higher pressure of 20 psig was chosen for interface resin transfer since it had to travel

further and line pressure drop was anticipated to be greater. A low degree of mixing was

noticed in the interface and cation layer during resin transfer. In both cases, using a low

backflow through the cationic bed during transfer helped stabilize the bed. In the

interface column there was about 6 inches of cation resin with a total bed height of 8

inches. These heights are primarily a reflection of the location of the interface levels

between the sampling/transfer points in the service column.

Two experiments were performed to quantify the effectiveness of the transfer process and

determine better conditions for transfer of both resins. The resin bed was initially air

mixed for 10 minutes. Gravity separation of the bed was done at 13.5 gpm and 100% bed

expansion. The anionic fraction in cation layer was measured after five backwashes. The

47

volumes of anion and heel resin transferred were measured. Anionic fraction in cation

layer was also measured after transfer to ascertain the change in cross contamination

during transfer of heel resin. For the first experiment, anion layer was transferred at 20

psi and interface at 30 psig and for the second experiment, anion layer was transferred at

30 psig and interface at 40 psig. After each experiment, both layers were mixed in the

resin storage vessel and transferred back to the cation regeneration vessel.

The results for both experiments are compared in Table 2. The purity of the cation layer

is not affected by the transfer process. Slightly lower values of anionic fraction is

observed due to drain and refilling of column after anion transfer during when more

anion resin is pushed towards the interface. In both experiments the pressures were

sufficient for effective resin transfer, however, increased pressure (experiment 2) gives

the operator better control of the transfer process. Hence there exists a large amount of

operator variability and relatively loose correlation between pressure and purity.

Table 2. Resin transfer study

Experiment 1 (20 & 30 psig)

Experiment 2 (30 & 40 psig)

Volume % of Anion layer transferred = 92.3 91.1

Volume % cation resin left behind = 79.2 77.9

Volume % comprised by heel resin = 16.2 17.2

Anionic fraction in cation layer before transfer = 0.015 0.019

Anionic fraction in cation layer after transfer = 0.014 0.010

48

DISCUSSION

The flowrates used were compared with the backwash expansion data given by Dowex

for both the resins as shown in Figure 40 (DOWEX 650C and 550A MSDS). The lines in

the graph indicate the percentage expansion that can be achieved for a set flowrate. This

helps to determine the flowrate needed to rinse a homogenous bed to remove fines

without losing any resin from the top of the column.

A summary of all the backwash conditions used are shown in Table 3. Though it was

easy to operate at the Dow cationic recommended curve, the bed could not be fluidized at

the Dow anionic conditions because of the presence, and higher density, of the cationic

resin in the mixed bed.

Figure 40. Comparison of conditions used and DOWEX expansion data

49

Table 3. Summary of backwash conditions used

S no.

Flowrate (gpm)

Flowrate (gpm/ft2)

Percent expansion

1 4.0 5.1 50 2 6.5 8.3 75 3 13.5 17.2 75 4 6.5 8.3 100 5 13.5 17.2 100 6 27.0 33.1 100 7 13.5 17.2 150 8 27.0 33.1 150

The effect of flow rate on mixing of resin beads during backwash could be visually

noticed and without sufficient expansion, results in increased cross contamination.

Reynolds numbers of the flow rates used can indicate the degree of viscous effects

experienced by the resin beads.

Table 4 shows the column Reynolds numbers corresponding to the flow rates used for the

backwash experiments. 4 and 6.5 gpm flows lie within the laminar region whereas flow

rates of 13.5 gpm and 27 gpm lie in the transition and turbulent regions respectively.

High Reynolds numbers indicate that the viscous effects acting on the resin beads are

greater. This causes increased mixing as the bed expands. No analysis of particle

Reynolds numbers in the bed were made.

Table 4. Estimation of column Reynolds numbers from flow rates

Flow rate (gpm) Velocity

(m/s) Reynolds number

4 0.0035 1050

6.5 0.0056 1710

13.5 0.0117 3538

27 0.0233 7061

50

Out of the 64 samples from the second backwash or later in all experiments for H/OH

form resin, only one data point exceeded 1% cross contamination. For the first

backwashes, only once did the cationic resin in anion layer exceed 1%. The lowest flow

rate required to achieve 100% expansion was found to be 6.5 gpm and 13.5 gpm was the

minimum flowrate required to achieve 150% expansion. Na/SO4 form resin backwash

results indicate that increased backwash time is required for better separation.

A working model can be developed to predict the separation and redistribution of the

resin beads with each backwash. The first step towards realization of such a model would

be to develop equations to predict the bead rise and fall characteristics. While developing

a force balance around a single resin bead, the forces to be taken into consideration are

the drag force FD, force due to buoyancy FB, force acting on an accelerating body FA and

gravitational force due to the weight W of the bead. The directions of forces acting on a

spherical bead during rise and fall are shown in figure 41.

Figure 41. Force balances for rise and fall of a resin bead

51

Equations (1) and (2) show the change in velocity over time for rise and fall of a single

resin bead. The derivations for both equations are presented in Appendix C.

Bead rise:

( | | ) ( )

Where,

, ( ) ,

,

Bead fall:

( ( ) |( )| ) ( )

Where,

, ( ) ,

,

52

CONCLUSION AND RECOMMENDATIONS

The following conclusions are drawn from this study:

Higher flowrate causes increased mixing due to turbulence whereas lower

flowrate gives significantly less cross contamination.

Greater bed expansion gives better separation at the same flow rate.

Increased water temperature leads to better separation of anion resin from cation

layer and decrease in cross contamination in anion layer with each backwash.

Interface isolation method is a viable resin transfer technique at pressures of 30/40

psig enabling better control during operation. However, loose correlation is

observed between pressure and purity since there is a large amount of operator

variability.

Possible future work includes defining different flow and expansion conditions for each

of the five backwashes to achieve low cross contamination. It is possible that once the

bed has a first backwash and a primary separation, that the next backwash could be

performed at a somewhat lower flowrate. The top half of the bed would still expand

while providing less turbulence and mixing at the interface. If the primary concern is to

liberate more cationic beads from the anionic resin this method may have some

advantage. While if the concern is the release of anionic beads from the cationic resin

then there must be sufficient flow to expand the cationic resin regardless.

The impact of steady state expansion from a high to low flow rate in a single backwash

also needs to be addressed. The water property with the largest variation over the range of

53

potential temperatures is viscosity. The impact of this factor can be evaluated in future

studies for different backwash conditions.

54

REFERENCES

Applebaum, S. B. (1968). Demineralization by ion exchange, Academic press. p. 173-182

Auerswald, D. C. (1983)."Two bed (cation-mixed bed) Polisher System San Onofre

Nuclear Generating Station". EPRI Condensate Polishing workshop, Los Angeles,

CA. p. 13-14

Auerswald, D. C. (1986)."Design and Performance of the Condensate Polishing System

at San Onofre Nuclear Generating Station". SCI meeting on Ultrapure Water,

University of Bristol, United Kingdom. p. 21

Auerswald, D. C. (1987)."San Onofre Nuclear Generating Station Polisher System:

Performance and Impact on Steam Generator Chemistry". EPRI Condensate

Polishing workshop, Chicago, IL. p. 45-46

Auerswald, D. C. and F. M. Cutler (1991)."Amsep: A New Approach for Mixed Bed

Resin Separation, Cleaning and regeneration". EPRI Workshop on Condensate

Polishing and Water Purification, Scottsdale, AZ. p. 26

Bevan, C. A. (1989)."Evaluating Deep Bed Condensate Polisher Performance for PWRs

using Morpholine". EPRI Condensate Polishing conference, Pittsburg, PA. p.

168-169

Chonde, Y. (1985). Interface beads for mixed bed ion exchange resin. US Pat 4,537,911

Crone, L. E. (1987)."Optimization of Deep Bed Condensate Polishers in the Presence of

Ammoniated Feedwater". EPRI Condensate Polishing Workshop, Chicago, IL. p.

134-136

Dorfner, K. (1972). Ion Exchangers: Properties and applications, Ann Harbor Science. p.

158-159

Emmet, J. R. (1977)."Condensate Polishing- A New Approach". BNES International

Conference on the Water Chemistry of Nuclear Reactor Systems, Bournemouth.

p. 112-115

Emmet, J. R. and P. M. Grainger (1979)."Ion exchange mechanisms in condensate

polishing". 40th International Water Conference, Pittsburg, PA. Engineers society

of western Pennsylvania. p. 91

55

EPRI (1997). Condensate Polishing Guidelines for PWR and BWR plants. Palo Alto, CA.

#TR101942R1. p. 221

Grimshaw, R. W. and C. E. Harland (1975). Ion-Exchange: Introduction to theory and

practice. London, The chemical society. p. 39-42.

Hattori, T. (1989)."Condensate Polishing with Boric Acid Treatment in Japanese PWR's".

EPRI Condensate Polishing Workshop, Little Rock, AK. p. 71-72

Kusano, H. and K. Nawata (1987)."Condensate Polishing in Japanese PWR's". EPRI

Condensate Polishing Workshop, Chicago, IL. p. 131

Lefevre, L. J. and T. Sato (1981). Separation of anion and cation resin in a mixed resin

bed. US Pat 4,264,439

Martin, D. (1952). The Industrial Chemist, McGraw-Hill. p. 28, 448

McCarthy, F. and G. O‟Connor (1992)."Ammonium Form Operation of Condensate

Polishing Plant for Long Periods in High pH systems at Moneypoint Power

Station". 52nd International Water Conference, Pittsburg, PA. Engineers Society

of Western Pennsylvania. p. 16

McGarvey, F. X. and R. Kunin (1951). Mixed Bed Deionization. US Pat 2,578,937

McGarvey, F. X. and R. Kunin (1954). Deionization of fluids. US Pat 2,692,244

Meyers, P. S. (1993)."The Amsep Process at Mohave Generating Station". EPRI

Condensate Polishing Workshop, New Orleans, LA. p. 6-7

Olijar, L. L. and E. Salem (1970)."Seprex: A New Technique in Ammoniated Cycle

Condensate Treatment". International Water Conference, Pittsburg, PA. p. 8-11

Osei-Gyimah, P. (1988). Inert separation beads for regeneration of mixed bed ion

exchange resins. US Pat 4,745,134

Sadler, M. A. (1986). Condensate Polishers for Brackish Water Cooled PWR's. NP-4550,

EPRI. NP-4550. p. 45-47

Sadler, M. A. (1993)."Developments in the production and control of Ultrapure water",

Cambridge, UK. The Royal Society of Chemistry. p. 12-14

Sadler, M. A. (1995). Method of regenerating resin beads for use in water purification.

US Pat 5,391,301

Sadler, M. A. (2001). Condensate Polishing Guidelines: Ammonium Form Operation.

Palo Alto, CA, EPRI. p. 122-123

56

Sadler, M. A. and H. R. Bolton (1996)."Exploring the Limits of Condensate Polishing".

EPRI Workshop on Water Purification and Condensate Polishing in the Steam

Cycle, San Antonio, Texas. p. 9

Salem, E. and C. C. Scheerer (1983)."A Unique Advance in Condensate Polishing at

Central Illinois Public Service Company". 44th International Water Conference,

Pittsburg, PA. p. 14,16

Shields, K., F. M. Cutler and M. A. Sadler (2006). Condensate Polishing State of

Knowledge Assessment. Palo Alto, CA, EPRI. p.

Stengel, A. and R. C. Pillow (1990)."Condensate Polishing Operating Experience 1987-

1990: Surry Power Station". EPRI Amine Workshop, Tampa, FL. p. 21

Townsend, R. P. (1993). Fundamentals of Ion Exchange. Cambridge, UK, The Royal

Society of Chemistry. p.

Webb, L. C. and B. A. Larkin (1996). Condensate Polishing Guidelines. Palo Alto, CA,

EPRI. p. 56-57

Yi, J. and G. L. Foutch (2004). "True multi-component mixed-bed ion-exchange

modeling." Reactive and Functional Polymers 60: 121-135.

57

APPPENDIX A

RESIN SEPARATION PILOT PLANT – OPERATING MANUAL

INTRODUCTION

The purpose of this operating manual is to provide all necessary information for the user

to effectively conduct resin separation studies on the pilot plant. Resin separation is the

most important step in the resin regeneration process. In mixed bed ion exchange, cation

and anion resin are present side by side, thoroughly mixed. Ineffective separation prior to

regeneration results in fully loaded cross contaminated beads with counterions and

subsequently increasing leakage of ions into the water. The topics covered are as follows,

Resin separation efficiency by gravity separation technique

Cross contamination measurement after separation

Resin transfer by interface separation method

EQUIPMENT SETUP

The assembled pilot plant (Figure 28) is comprised of the following units:

DILUTION/ DEMIN WATER RESIN SLUICE PUMP

CAPACITY: 4 m3/ h

DISCHARGE PRESSURE: 50 PSIG

MOTOR: 3 phase

58

Figure 1. Resin separation Pilot plant (Aquatech International Corp.)

CATION REGENERATION/SEPARATION VESSEL (CR)

DIAMETER: 304.8 mm

DESIGN PRESSURE: 75 PSIG

MATERIAL: PLASTIC or FRP or CS Epoxy Lined

STRAIGHT HEIGHT: 1500 mm

59

ANION REGENERATION VESSEL (AR)

DIAMETER: 304.8 mm




RESIN STORAGE VESSEL (RS)

DIAMETER: 304.8 mm




DI WATER STORAGE TANK

MATERIAL: PPL or equivalent

RESIN USED

CATION RESIN: DOWEX 650C(H)

ANION RESIN: DOWEX 550A(OH)

SAFETY AND ENVIRONMENTAL ISSUES

Water should be kept away from the computers and electrical equipment to avoid electric

shock and electrocution. Any spilled water should be immediately dried to avoid slipping.

Use caution to prevent tripping in the work area. Use caution while climbing the ladder to

reach the top of the tanks. Always be aware of the surroundings and wear safety glasses

at all times. The water used for this testing is mostly recycled and the pure water obtained

may be used for other experiments. The resins are not hazardous to health. Resin spilled

60

on the floor however is a slip hazard. Avoid skin contact with sodium hydroxide,

sulphuric acid and ammonia (Material Safety Data Sheets).

Utmost care is required while handling the valves and piping since increased physical

strain can lead to breakage. In case the resins are regenerated, the sodium hydroxide and

sulphuric acid used as regenerants are mixed in a separate tank after completion of the

process to bring pH back to acceptable levels before disposal.

TEST PROCEDURE – OVERVIEW

The three vessels in the pilot plant as shown in Figure 1 are:

Cation Regeneration (CR) vessel

Anion Regeneration (AN) vessel

Resin Storage (RS) vessel

Cation resin and anion resins are separated in the Cation Regeneration / Separation vessel

(CR) by backwashing and by gravity, cation resin being heavier will settle below the

anion resin component. An air scrub/scour is used to mix resins and create a homogenous

bed prior to separation. Gravity separation of the two resins is performed using step-by-

step backwash runs. This is performed by fixing a backwash flow rate and resin bed

expansion height (e.g. 100% resin expansion, 75% resin expansion etc.) To determine a

specific resin expansion, refer to resin manufacturer‟s specification (resin expansion rate

vs. backwash flow linear velocity) and the required backwash flow rate will be

determined considering the cross sectional area of the test separation vessel.

During backwashing steps the DI water enters from the bottom of the vessel containing

resins in an upward direction thereby fluidizing the resin bed. Although there is a sharp

61

difference between in resin particle densities and particle size distribution, even after

gravity separation there is a level of intermixing at the interface. Separation of the resin is

directly proportional to degree of fluidization of the mixed resin. Degree of fluidization is

affected by the backwash velocity of water and duration of backwash. The anion layer on

top is transferred to the Anion Regeneration (AN) vessel and the interfacial layer is

transferred to the Resin Storage (RS) vessel.

PRE-OPERATION CHECKS

Before operating the pilot plant, ensure that:

Water level in the DI tank is adequate

Instruments in the pilot plant are in good working condition

All leaks are fixed

Availability of air is checked

WATER FILLING AND HYDRO TESTING

The filling and hydro testing process is done one tank at a time. Check that all valves are

closed before beginning the test.

CR VESSEL:

Open the valve H1 (Lab‟s RO water inlet) to fill the DI water tank if the tank is not

filled. Ensure the tank is filled no closer than a foot from the top.

Open valve DI1 and DI4. Check for leaks around the pump.

Partially open (¼ turn) the recycle valve DI6. Refer figure 29 for valve locations.

Now open valves CR1 (inlet valve to CR tank) and the vent valve CR2 on the top.

62

Turn on the DI pump. Check PID-1 and flow indicator FL2. Ensure pressure is not

more than 55 psig while the CR tank is getting filled.

Now close valve DI6 completely.

During the fill check for any leaks.

Fill CR vessel until water drains out from the vent and then close valve CR2.

Let the pressure rise up to 55 psig. Turn off the pump and look for pressure decay.

Wait for ten minutes. Pressure drop of about 2-3psi is admissible.

Now close valve CR1.

Figure 2. Inlet and recycle valves

63

AN VESSEL:

Keep valve DI6 partially open (¼ turn).

Open inlet valve AN1 and vent valve AN2.

Turn on the DI pump and set pump speed to level 3. Check PID-1 and flow indicator

FL2. Ensure pressure is not more than 55 psig while the AN vessel is getting filled.


Check for leaks during the filling procedure.

Fill the vessel until water drains out from the vent and close vent valve AN2.

Let pressure increase to 55 psig then turn off the pump and look for pressure decay.


Now close the valve AN1.

RS VESSEL:

Keep valve DI6 partially open (¼ turn).

Open valve RS1 and vent valve RS2.

Turn on the DI pump and set pump speed to level 3. Check PID-1 and flow indicator

FL2. Ensure pressure is not more than 55 psig while the AN vessel is getting filled.


Check for leaks during the filling procedure.

Fill the tank until water drains out from the vent and close the vent valve RS2.

Let pressure increase to 55 psig then turn off the pump and look for pressure decay.


Now close the valve RS1.

64

RECYCLING WATER TO DI TANK FROM VESSELS

In all three vessels water can be drained back to the DI tank by gravity and by

pressurising vessels with air. Ensure the pressure gage for air supply inlet is at 10 psig

and the air inlet valve A1 is open.

CR VESSEL:

Open CR vessel bottom valve CR14 and drain valve D1.

Close vent valve CR2.

Open air valve A2 to begin building pressure within CR vessel which will force water

to flow back in to the DI tank.

Close and open air valve A2 occasionally to regulate water flow.

Once all the water has been recycled, close valve CR14 and then open D1.

Open vent valve CR2 slowly to remove air from the vessel and decrease pressure.

AN & CR VESSELS:

Open drain valve D1. Let AN vessel bottom valve, AN13 and vent valve AN2 be

closed.

Open air valve AN10 in the bottom to start bubbling air into AN vessel until pressure

in the vessel is the same as the inlet air pressure.

Now open bottom valve AN13 to let the excess air push the water back into the DI

tank.

Repeat the above steps till all water is recycled from AN vessel.

After the recycling is complete open vent valve AN2.

65

The process is similar for RS vessel where RS2 is the vent valve, RS12 is the drain

valve and RS8 is the air inlet valve.

RESIN LOADING

A specific amount of cation and anion resin can be loaded into the Resin Separation

vessel. The cation resin is filled first followed by the anion resin. Make sure the DI tank

is full before beginning this procedure.

Open bottom valve CR14 and drain valve D1. Close vent valve CR2 and open Air

inlet A2 to facilitate effective recycling of water back into the DI tank.

Once water level comes down to about a foot from the bottom of the vessel, close

valve A2 and slowly open vent valve CR2 to remove excess air from the CR vessel.

Connect the transfer hose leading from the cation drum to valve E1 which in turn

leads to the resin eductor. Ensure E1 is closed initially.

Open valve R1 (present beneath the DI tank). Close the bottom valve CR14 and drain

valve D1.

Keep valve DI6 partially open (¼ turn). Turn on the pump and increase to maximum

speed.

Check FI1 to note flow and then close valve DI6.

Open valve E1 to create suction. Now insert the transfer hose into the resin drum

while stirring the resin gently.

Turn off the pump once desired amount is transferred. Wait for excess resin to return

back to the resin drum and close valve E1.

By visual observation, allow time for the resin to settle.

66

If desired amount of resin was not transferred, drain the CR vessel and repeat the

process.

Connect Valve E1 to anion drum and repeat the process to transfer required amount

of anion resin to the CR vessel.

RESIN SEPARATION

The cation and anion resin beads are separated by virtue of difference in their densities.

This separation is achieved during the backwash and settling procedure where the less

dense anion resin forms a top layer and the slightly heavier cation resin forms the bottom

layer. Before separation the resin bed has to be air mixed to form a homogenous mixture.

AIR MIXING:

Ensure that the water level is only an inch above the resin bed prior to air mixing. Open

the air inlet valve CR11 and vent valve CR2.

Open the air inlet valve A1 and maintain a flow rate of at least 4 m3/hr.

Carryout the air mixing process for 10 minutes.

Close valves A1 and CR11.

Allow two minutes for the resin mixture to settle.

BACKWASH & SETTLING STEPS:

Open valves CR14 and vent valve CR2. Let valve DI6 be partially open and ensure

all other valves are closed.

Turn on the pump while set at minimum speed and check FI1 to note flow.

Close valve DI6 and adjust pump speed to obtain desired flowrate.

67

The anion and cation resins separate during the back wash.

Continue backwash till bed reaches the desired level of expansion.

Stop the pump. Close bottom valve CR14 and allow time for resin to settle.

The anion resin should now be on top and cation resin in the bottom.

The interface consists of a mixture of both anion and cation resins. It is called the

„heel resin‟.

Drain water back as shown in section 4.4 and repeat backwash if necessary.

MEASURING CROSS CONTAMINATION

Once resin separation is completed efficiency of separation can be measured by drawing

samples from the take-off ports (figure 30).

Recycle all water from the CR vessel to the DI tank.

Open the chosen take-off port and remove the port insert.

Insert a glass tube through the port to remove approximately 5-10 grams of resin.

Transfer sample to a tall glass vial containing 20% NaCl solution. Allow resins to

settle.

The NaCl will allow the resins to separate due to difference in density.

Decant or siphon the less dense anion beads from the top layer of the solution and

place on filter paper. Allow excess solution to drain from the beads.

Remove the cation resin beads from the remaining mixture using a sieve or filter and

let dry.

Wash both sets of resin beads with water to dissolve any salt residue and oven dry

the samples overnight at 35oC.

68

Once dry, weigh the filter paper containing the resin. Remove resin from the filter

paper and weigh filter paper separately.

The weight of each resin can now be determined by subtracting the weight of the

filter paper from the weight of the resin and filter paper combination.

Record weights of anion and cation resins and calculate relative weight fractions.

Figure 3. Resin transfer and sampling take-off ports

RESIN TRANSFER

In this step the anion resin is transferred to the AN vessel and the heel resin is transferred

to the RS vessel through the take-off ports as shown in Figure 30. Ensure all valves are

closed before beginning the procedure.

69

ANION RESIN TRANSFER TO AN VESSEL:

Connect the chosen take-off port to AN5 using the resin transfer tube. Ensure both

valves are closed.

Open the vent valves AN2 and CR2.

Open the valve DI6 partially (¼ turn) and turn on the pump while set at minimum

speed.

Note the flow FI1 and open valve CR1 to start filling the CR vessel. Close valve DI6.

Once water flows out of the vent, close CR2 and allow CR vessel to pressurize up to

25 psig.

To begin resin transfer open the selected take-off valve and then open valves AN5

and AN8.

Open the bottom valve CR14 to stabilize the bed during transfer.

Once the transfer is complete open valve DI6. Close valves AN5 and AN8.

Turn off the pump and close valves CR1, CR14 and DI6.

Wash and collect the resin remaining in the resin transfer tube separately.

HEEL RESIN TRANSFER TO RS VESSEL:

Select an appropriate take-off port from CR vessel (CR5 to CR10) for transferring the

heel resin to RS vessel.

1. Connect the resin transfer tube from chosen port to RS13. Ensure both valves are

closed.

2. Open vent valves CR2 and RS2.

3. Open valve DI6 partially (¼ turn) and turn on the pump while set at minimum speed.

70

4. Note the flow FI1. Open valve CR1 and close valve DI6.

5. Once water flows out of the vent close CR2 and allow CR vessel to pressurize up to

25 psig.

6. Open selected take-off valve, and RS13 leading to the RS vessel.

7. Open bottom valve CR14 to stabilize bed during transfer.

8. After Transfer is complete open valve DI6. Close the take-off valve and RS13.

9. Close DI4, CR1 and CR14 to stop supply of water to CR vessel.

10. Turn off the pump and close the valve DI6.

RESIN TRANSFER TO CR VESSEL FROM AN AND RS VESSELS:

Transfer of resin to CR vessel from AN and RS vessel is done in two steps. The anion

resin in the AN vessel is first transferred to the RS vessel. The resin from the RS vessel is

then transferred to the CR vessel.

1) Transfer from AN vessel to RS vessel

Open valves AN1, vent valves AN2 and RS2. Keep valve DI6 partially open (¼

turn).

Start pump to begin filling the AN vessel with water at a flowrate of 8 gpm.

Close valve DI6 and close off vent valve AN2 as soon as water flows out.

Let the pressure in AN vessel increase up to 25 psig.

Now open AN9 and RS6 to begin transfer. Open bottom valve AN13 to lift the bed

while transfer takes place.

The transfer process cannot be completed in one shot. Once the RS vessel fills up,

close RS6 and AN9. Open DI6 and turn the pump off.

71

Drain water from RS vessel and repeat the transfer process till most of the resin has

been transferred.

The resin remaining in the bottom can be transferred though the bottom resin transfer

pipe.

Pressurize AN vessel again up to 25 psig. Open valves AN14 and RS6 to transfer the

remaining resin.

Repeat above step to ensure complete transfer.

Drain water from AN vessel to the DI tank.

2) Transfer from RS vessel to CR vessel

Open valves RS1, vent valves RS2 and CR2. Keep valve DI6 partially open (¼ turn).

Start pump and begin filling RS vessel water at a normal flowrate.

Close valve DI6 and vent valve RS2 as soon as water flows out.

Let the pressure in RS vessel increase up to 25 psig.

Open valve RS7 to start transfer and open bottom valve RS12 to slowly lift the bed

while transfer takes place.

When the CR vessel reaches maximum water capacity, drain water from the CR

vessel and repeat steps 4 and 5 until 50% of the resin is transferred.

To transfer the bottom portion of the resin the resin can be transferred through the

bottom valve RS9.

Connect the resin transfer hose from the topmost take-off port to AN5.

Drain water completely from the CR vessel before beginning the transfer.

Pressurize RS vessel to 25 psig and open valves RS9, AN5 and the take-off port valve

to transfer the resin.

72

Drain CR vessel and repeat above step to ensure complete transfer of resin.

Recycle excess water back to the feed tank and close off all valves after the transfer is

complete.

RESIN REGENERATION

After gravity separation, transfer the anion resin to the AN vessel and the heel resin to

the RS vessel.

Based on the total gram equivalents of anion and cation resin present in the AN and

CR vessels respectively, caustic and sulphuric acid solutions containing the same

gram equivalents of OH- and H

+ ions have to be prepared.

Transfer the prepared solutions in to the columns using a chemical pump. For CR

vessel, the top most take-off port can be used as an inlet for the sulphuric acid

solution. For AN vessel, there exists an auxiliary inlet at the top of the vessel for

transferring the caustic solution.

Let the resins soak for 45 minutes after which the sulphuric and caustic solutions

should be drained safely to the storage tank in the basement. This followed by rinsing

the resins with DI water which is also sent to the storage stank.

Fill the CR vessel completely with water and slowly rinse the bed while measuring

the conductivity of the exiting the water stream using the conductivity sensor. After

the conductivity reaches a sufficiently low value, the water can be recycled back to

the DI tank during the rinse process. The same process is repeated for the AN vessel.

Ensure that the solution in storage tank is neutralised before it is dumped.

73

APPPENDIX B

EXPERIMENTAL DATA

INTRODUCTION

The following experimental data shows change in position of the interface with each

backwash for different experimental conditions. Interfacial height gives a visual

perception of how well the anion and cation resin separate after each backwash. The

tables also show the anion and cation sampling distance relative to the interface. Other

details mentioned are the air mix conditions and the total bed height.

H/OH FORM

Experiment 1: 13.5 gpm and 100% bed expansion

Air mixing Bed Conditions

Rate

(m3/h)

Time

(min)

Cation sampling height

(cm)

Anion sampling height

(cm)

Bed

height

(cm)

3.5 10 25 54 77

Runs

Interface

height

(cm)

Sampling distance from interface (cm)

Cation resin anion resin

1 59.0 18.0 13.0

2 52.0 11.0 20.0

3 50.0 9.0 22.0

4 49.5 8.5 22.5

5 49.4 8.4 22.6

74



Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 76

Runs

Interface

height

(cm)



1 49.0 8.0 23.0

2 47.0 6.0 25.0

3 46.0 5.0 26.0

4 46.0 5.0 26.0

5 46.0 5.0 26.0



Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 76

Runs

Interface

height

(cm)



1 57.0 16.0 15.0

2 50.5 9.5 21.5

3 49.0 8.0 23.0

4 48.5 7.5 23.5

5 48.5 7.5 23.5

75



Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 77

Runs

Interface

height

(cm)



1 50.5 9.5 21.5

2 49.2 8.2 22.8

3 49.0 8.0 23.0

4 48.8 7.8 23.2

5 48.5 7.5 23.5



Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 75

Runs

Interface

height

(cm)



1 51.5 10.5 20.5

2 49.0 8.0 23.0

3 48.5 7.5 23.5

4 48.0 7.0 24.0

5 48.0 7.0 24.0

76

Experiment 6: 4 gpm and 50% bed expansion


Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 76

Runs

Interface

height

(cm)



1 49.5 8.5 22.5

2 47.5 6.5 24.5

3 47.0 6.0 25.0

4 46.5 5.5 25.5

5 46.0 5.0 26.0

Experiment 7: 6.5 gpm and 75% bed expansion (40οC)


Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 75

Runs

Interface

height

(cm)



1 51.0 10.0 21.0

2 48.5 7.5 23.5

3 47.5 6.5 24.5

4 47.0 6.0 25.0

5 47.0 6.0 25.0

77

Na/SO4 FORM



Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 59.5

Runs

Interface

height

(cm)



1 43.5 18.5 10.5

2 39.5 14.5 14.5

3 39.5 14.5 14.5

4 38.5 13.5 15.5

5 38.0 13 16



Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 59.5

Runs

Interface

height

(cm)



1 42.5 17.5 11.5

2 39.0 14.0 15.0

3 38.0 13.0 16.0

4 38.0 13.0 16.0

5 37.5 12.5 16.5

78



Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 59.5

Runs

Interface

height

(cm)



1 40.5 15.5 13.5

2 37.5 12.0 16.5

3 37.0 12.0 17.0

4 37.0 12.0 17.0

5 36.5 11.5 17.5

Experiment 4: 4 gpm and 50% bed expansion


Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 59

Runs

Interface

height

(cm)



1 39.5 14.5 14.5

2 37.5 12.5 16.5

3 37.0 12.0 17.0

4 37.0 12.0 17.0

5 37.0 12.0 17.0

79



Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 58

Runs

Interface

height

(cm)



1 38.0 13.0 16.0

2 37.0 12.0 17.0

3 36.5 11.5 17.5

4 36.5 11.5 17.5

5 36.3 11.3 17.5



Rate

(m3/h)

Time

(min)


(cm)


(cm)

Bed

height

(cm)

3.5 10 25 54 58

Runs

Interface

height

(cm)



1 37.0 12.0 17.0

2 36.7 11.7 17.3

3 36.0 11.0 18.0

4 36.0 11.0 18.0

5 36.0 11.0 18.0

80

APPPENDIX C

BEAD RISE AND FALL MODEL

The following forces need to be taken into consideration to model the rise and fall of a

spherical resin bead in water:

1. The gravitational pull, W = mg

2. The buoyant force FB, which is equal to the weight of the weight of the fluid

displaced by the body. It has the expression FB = mf g

3. The force on an accelerating body, FA, which is the kinetic energy associated with

moving the resin bead against a drag force. It has the expression FA =

4. The drag force, FD, caused by the fluid viscosity given by

| |

BEAD SETTLING

A single spherical particle is considered to fall freely along the z axis in the direction of

gravitational acceleration. Wall effects and particle interactions are neglected.

Figure 1. Forces acting on a settling spherical bead

82

BEAD RISE

Neglecting the wall effects and particle interactions, a single spherical particle is

considered to rise along the z-axis.

Figure 2. Forces acting on a rising spherical bead

From Newton‟s law of motion,

( )|( )|

(

)

( )

( )|( )|

( )

( )

( )

|( )|

( )

Substituting,

and

( )

( )

( )|( )|

( )

83

Rearranging,

( ( ) |( )| )

Where,

, ( ) ,

,

VITA

Bharathwaj Gopalakrishnan

Candidate for the Degree of

Master of Science

Thesis: SEPARATION OF RESIN TYPES IN MIXED BED ION EXCHANGE

COLUMNS

Major Field: Chemical Engineering

Biographical:

Education:

Completed the requirements for the Master of Science in Chemical Engineering

at Oklahoma State University, Stillwater, Oklahoma in May 2011

Completed the requirements for the Bachelor of Science in chemical

engineering at St. Joseph‟s College of Engineering (Affiliated to Anna

University), Chennai, Tami Nadu, India in 2009.

Experience:

Graduate Teaching and Research Assistant at the School of Chemical

Engineering, Oklahoma State University, Aug., 2009, to May, 2011.

.

ADVISER‟S APPROVAL: Dr. Gary L. Foutch

Name: Bharathwaj Gopalakrishnan Date of Degree: May 2011

Institution: Oklahoma State University Location: Stillwater, Oklahoma

Title of Study: SEPARATION OF RESIN TYPES IN MIXED BED ION EXCHANGE

Pages in Study: 83 Candidate for the Degree of Master of Science

Major Field: Chemical Engineering

Scope and Method of Study: Resin separation is a vital step prior to regeneration of Ion

Exchange resins. The degree of separation determines the extent of sodium and

sulfate or chloride into the system from caustic and acid regenerants, respectively.

This study is aimed at determining the backwash conditions that would yield the

best separation and also investigate the efficiency of interface isolation method

for resin transfer.

Findings and Conclusions: Flowrate and height of bed expansion are the two important

parameters controlling resin separation. Increased backwash time along with low

flow leads to better separation of a mixed bed. Interface isolation method is a

viable resin transfer technique. The two factors affecting transfer are pressure and

operator skill

gopalakrishnan okstseparation of resin types in mixed bed ion exchange columnate 0664m 11405

Documents

separation of resin

effective resin separation

resin separation techniques

density separation

conesep separation system

fullsep separation system

graduate college

mixed beds